Hedonic Pricing Models

Sawyer Benson’s Master Thesis

Janurary 10, 2022

1. Model Design: Checks & Corrections

1.1 Accounting for Heteroskedasticity

# All-inclusive model
lm_pre_alpha <- lm(sold_price ~ . , data = data_factor_core)
summ(lm_pre_alpha)

# pre_alphaing for heteroskedasticity
#  a. Graphically
par(mfrow = c(2,2))
plot(lm_pre_alpha)

#autoplot(lm_pre_alpha)

#  b. Statistically
ols_test_breusch_pagan(lm_pre_alpha) # Breusch-Pagan test

# - Resolving Heteroskedasticity using heteroskedasticity-consistent (HC) variance covariance matrix

# Compare models
stargazer(lm_pre_alpha,
          coeftest(lm_pre_alpha, vcov = vcovHC(lm_pre_alpha, method = "White2", type = "HC0")),
          coeftest(lm_pre_alpha, vcov = vcovHC(lm_pre_alpha, method = "White2", type = "HC1")),
          type = "text")

1.2 Accounting for Interactions

Note: Advisor suggested not to inlude interaction terms except for specific testing.

1.3 Accounting for Non-linearity

1.3.1 Age

# Age
a <- ggplot(data_factor, aes(x = age , y = sold_price)) +
     geom_smooth(aes(fill = infections_period)) +
     geom_smooth(linetype = "dashed", color = "grey32") +
     theme_minimal() +
     #scale_fill_manual(values=c(very_low, med)) +
     labs(title = "Age and Price",
          x = "Age",
          y = "Price") +
     scale_fill_manual(values = c(very_low, med), 
                      name = "Infection Period",
                      labels = c("Pre", "Post"))
     

a

`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'
`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'

# Actual vs. fit

# Model with non-linear addition
lm_pre_alpha_age <- lm(sold_price ~ . + I(age^2), data = data_factor_core)
summ(lm_pre_alpha_age)

MODEL INFO:
Observations: 24394 (18 missing obs. deleted)
Dependent Variable: sold_price
Type: OLS linear regression 

MODEL FIT:
F(66,24327) = 36748.75, p = 0.00
R² = 0.99
Adj. R² = 0.99 

Standard errors: OLS
-------------------------------------------------------------------------
                                           Est.      S.E.   t val.      p
----------------------------------- ----------- --------- -------- ------
(Intercept)                           -12663.15   9510.30    -1.33   0.18
property_typeDUP                       -1420.31   2871.50    -0.49   0.62
property_typeOTH                       -2648.78   2053.45    -1.29   0.20
property_typePAT                        -626.11    929.50    -0.67   0.50
property_typeSGL                        1784.81    437.71     4.08   0.00
property_typeTNH                         510.26    551.55     0.93   0.35
ac_typenone                              -83.17    380.75    -0.22   0.83
ac_typenot_central                     -1707.03    245.87    -6.94   0.00
list_price                                 0.98      0.00   888.10   0.00
patio1                                   775.22    126.90     6.11   0.00
school_general1                          151.15    161.81     0.93   0.35
photo_count                              -29.40      7.65    -3.84   0.00
pool1                                    -91.33    211.57    -0.43   0.67
roof_typeother                          1123.91    232.86     4.83   0.00
roof_typeshingle                        1815.51    262.56     6.91   0.00
roof_typeslate                           404.62   1113.88     0.36   0.72
gas_typenatural                         4180.15   8533.24     0.49   0.62
gas_typenone                            3729.55   8529.10     0.44   0.66
gas_typepropane                         -124.79   8729.36    -0.01   0.99
gas_typeunknown                         3388.93   8528.19     0.40   0.69
out_building1                           -424.03    137.78    -3.08   0.00
area_living                               -0.82      0.27    -3.01   0.00
land_acres                              -305.11    154.40    -1.98   0.05
appliances1                              850.31    172.71     4.92   0.00
garage1                                  623.51    127.04     4.91   0.00
property_conditionnew                  -4181.83    789.37    -5.30   0.00
property_conditionother                 -425.40    169.02    -2.52   0.01
energy_efficient1                        589.01    141.61     4.16   0.00
exterior_typemetal                       -78.23    402.23    -0.19   0.85
exterior_typeother                        35.68    167.52     0.21   0.83
exterior_typevinyl                       390.50    185.92     2.10   0.04
exterior_typewood                       -646.79    262.80    -2.46   0.01
exterior_featurescourtyard              2427.56   1466.46     1.66   0.10
exterior_featuresfence                  1028.53    614.35     1.67   0.09
exterior_featuresnone                   1539.28    615.57     2.50   0.01
exterior_featuresporch                   950.44    629.21     1.51   0.13
exterior_featurestennis_court            536.63   1724.79     0.31   0.76
fireplace1                               408.15    131.51     3.10   0.00
foundation_typeslab                     1016.20    191.35     5.31   0.00
foundation_typeunspecified              -110.61    229.01    -0.48   0.63
area_total                                -0.15      0.16    -0.97   0.33
beds_total1                             -441.27   3175.53    -0.14   0.89
beds_total2                             -837.36   3145.00    -0.27   0.79
beds_total3                             -195.13   3148.38    -0.06   0.95
beds_total4                              639.15   3154.44     0.20   0.84
beds_total5                             -183.17   3212.66    -0.06   0.95
bath_full1                              2051.07   3355.33     0.61   0.54
bath_full2                              2540.17   3355.08     0.76   0.45
bath_full3                              2065.05   3363.12     0.61   0.54
bath_full4                             -2648.80   3755.25    -0.71   0.48
bath_full6                             -5631.95   9199.84    -0.61   0.54
bath_half1                              -295.30    166.82    -1.77   0.08
bath_half2                             -1640.99   1098.85    -1.49   0.14
bath_half3                              1510.31   6029.59     0.25   0.80
bath_half4                              8533.97   8532.40     1.00   0.32
bath_half5                             -8590.00   4932.13    -1.74   0.08
age                                     -124.18     11.06   -11.23   0.00
dom                                       -7.97      1.08    -7.37   0.00
sold_date                                  0.17      0.07     2.64   0.01
sewer_typeseptic                        -185.34    237.12    -0.78   0.43
sewer_typeunspecified                    275.03    129.35     2.13   0.03
property_stylenot_mobile                2262.11    353.36     6.40   0.00
subdivision1                             396.83    151.52     2.62   0.01
water_typewell                           641.52    599.64     1.07   0.28
waterfront1                            -1671.69    225.43    -7.42   0.00
bottom25_dom1                           2367.79    158.88    14.90   0.00
I(age^2)                                   1.16      0.14     8.32   0.00
-------------------------------------------------------------------------

# Marginal effects data frames
ggpredict_1 <- ggpredict(lm_pre_alpha, terms = "age")
ggpredict_2 <- ggpredict(lm_pre_alpha_age, terms = "age")

# Plots
b <- ggplot(data_factor_core, aes( x = age)) +
   geom_smooth(data_factor_core, mapping = aes(y = sold_price), color = "grey50") +
   geom_smooth(ggpredict_1, mapping = aes(x, predicted), linetype = "dashed", color = very_low) +
   geom_smooth(ggpredict_2, mapping = aes(x, predicted), linetype = "dashed", color = med) +
     labs(title = "Age and Price",
          x = "Age",
          y = "Prediction")

# Look at age & age^2 alone to see impact on more relevant y-axis scale
c <- ggplot() +
   geom_smooth(ggpredict_1, mapping = aes(x, predicted), linetype = "dashed", color = very_low) +
   geom_smooth(ggpredict_2, mapping = aes(x, predicted), linetype = "dashed", color = med) +
     labs(title = "Age and Price",
          x = "Age",
          y = "Prediction") 

a

`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'
`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'

gridExtra::grid.arrange(b,c, nrow =2, ncol = 1)

`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'
`geom_smooth()` using method = 'loess' and formula 'y ~ x'
`geom_smooth()` using method = 'loess' and formula 'y ~ x'
`geom_smooth()` using method = 'loess' and formula 'y ~ x'
`geom_smooth()` using method = 'loess' and formula 'y ~ x'

1.3.2 Living Area

# Living Area

# General graphing
a <- ggplot(data_factor, aes(x = area_living , y = sold_price)) +
     geom_smooth(aes(fill = infections_period)) +
     geom_smooth(linetype = "dashed", color = "grey32") +
     theme_minimal() +
     #scale_fill_manual(values=c(very_low, med)) +
     labs(title = "Living Area and Price",
          x = "Living Area",
          y = "Price") +
     scale_fill_manual(values = c(very_low, med), 
                      name = "Infection Period",
                      labels = c("Pre", "Post"))

a

`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'
`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'

ggplot(data_factor, aes(x = area_living , y = sold_price/area_living)) + 
    geom_point(aes(color = infections_period), alpha = 0.15) + 
    geom_smooth(aes(color = infections_period)) +
    geom_smooth(color = "grey50", linetype = "dashed") +
    theme_minimal()

`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'
`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'

# Actual vs. fit
# Model with non-linear addition
lm_pre_alpha_area <- lm(sold_price ~ . + I(area_living^2), data = data_factor_core)
summ(lm_pre_alpha_area)

MODEL INFO:
Observations: 24394 (18 missing obs. deleted)
Dependent Variable: sold_price
Type: OLS linear regression 

MODEL FIT:
F(66,24327) = 36741.79, p = 0.00
R² = 0.99
Adj. R² = 0.99 

Standard errors: OLS
-------------------------------------------------------------------------
                                           Est.      S.E.   t val.      p
----------------------------------- ----------- --------- -------- ------
(Intercept)                           -21671.88   9522.17    -2.28   0.02
property_typeDUP                       -1333.48   2871.86    -0.46   0.64
property_typeOTH                       -2804.77   2053.59    -1.37   0.17
property_typePAT                        -620.44    929.59    -0.67   0.50
property_typeSGL                        1770.31    437.77     4.04   0.00
property_typeTNH                         370.43    551.95     0.67   0.50
ac_typenone                               62.25    381.06     0.16   0.87
ac_typenot_central                     -1498.05    246.37    -6.08   0.00
list_price                                 0.98      0.00   896.13   0.00
patio1                                   798.99    126.79     6.30   0.00
school_general1                          241.58    161.60     1.49   0.13
photo_count                              -34.70      7.62    -4.55   0.00
pool1                                    -73.45    211.70    -0.35   0.73
roof_typeother                          1098.57    233.01     4.71   0.00
roof_typeshingle                        1920.08    261.94     7.33   0.00
roof_typeslate                           536.02   1113.83     0.48   0.63
gas_typenatural                         4855.78   8534.04     0.57   0.57
gas_typenone                            4318.56   8530.00     0.51   0.61
gas_typepropane                           87.56   8730.21     0.01   0.99
gas_typeunknown                         3979.77   8529.01     0.47   0.64
out_building1                           -490.59    137.56    -3.57   0.00
area_living                                6.54      0.95     6.85   0.00
land_acres                              -285.71    154.41    -1.85   0.06
appliances1                              921.60    172.47     5.34   0.00
garage1                                  666.84    126.78     5.26   0.00
property_conditionnew                  -3617.20    784.80    -4.61   0.00
property_conditionother                 -364.93    168.83    -2.16   0.03
energy_efficient1                        601.45    141.63     4.25   0.00
exterior_typemetal                        16.32    402.32     0.04   0.97
exterior_typeother                        58.29    167.52     0.35   0.73
exterior_typevinyl                       417.26    185.92     2.24   0.02
exterior_typewood                       -554.23    262.89    -2.11   0.04
exterior_featurescourtyard              2805.14   1465.90     1.91   0.06
exterior_featuresfence                  1048.09    614.40     1.71   0.09
exterior_featuresnone                   1584.20    615.59     2.57   0.01
exterior_featuresporch                  1119.15    628.89     1.78   0.08
exterior_featurestennis_court            870.69   1724.92     0.50   0.61
fireplace1                               264.36    131.42     2.01   0.04
foundation_typeslab                      819.18    189.82     4.32   0.00
foundation_typeunspecified              -213.55    228.49    -0.93   0.35
area_total                                -0.27      0.16    -1.71   0.09
beds_total1                            -1072.82   3176.15    -0.34   0.74
beds_total2                            -2553.13   3149.94    -0.81   0.42
beds_total3                            -2327.60   3156.66    -0.74   0.46
beds_total4                            -1389.50   3161.99    -0.44   0.66
beds_total5                            -1954.03   3218.17    -0.61   0.54
bath_full1                              3642.81   3358.78     1.08   0.28
bath_full2                              3719.16   3356.69     1.11   0.27
bath_full3                              3740.67   3367.07     1.11   0.27
bath_full4                              -544.43   3761.07    -0.14   0.88
bath_full6                             -3367.38   9198.35    -0.37   0.71
bath_half1                              -274.20    167.01    -1.64   0.10
bath_half2                             -1480.96   1099.19    -1.35   0.18
bath_half3                              1451.22   6030.19     0.24   0.81
bath_half4                              7762.17   8533.71     0.91   0.36
bath_half5                             -8041.27   4933.05    -1.63   0.10
age                                      -37.00      3.75    -9.87   0.00
dom                                       -8.28      1.08    -7.66   0.00
sold_date                                  0.28      0.06     4.35   0.00
sewer_typeseptic                        -304.75    236.80    -1.29   0.20
sewer_typeunspecified                    258.97    129.37     2.00   0.05
property_stylenot_mobile                2105.89    353.77     5.95   0.00
subdivision1                             401.26    151.53     2.65   0.01
water_typewell                           557.86    599.60     0.93   0.35
waterfront1                            -1642.44    225.40    -7.29   0.00
bottom25_dom1                           2331.25    158.82    14.68   0.00
I(area_living^2)                          -0.00      0.00    -8.04   0.00
-------------------------------------------------------------------------

# Model with single-variable fit
lm_pre_alpha_area_single <- lm(sold_price ~ area_living, data = data_factor_core)
summ(lm_pre_alpha_area_single)

MODEL INFO:
Observations: 24412
Dependent Variable: sold_price
Type: OLS linear regression 

MODEL FIT:
F(1,24410) = 14244.19, p = 0.00
R² = 0.37
Adj. R² = 0.37 

Standard errors: OLS
-------------------------------------------------------
                         Est.      S.E.   t val.      p
----------------- ----------- --------- -------- ------
(Intercept)         -20238.66   1644.55   -12.31   0.00
area_living            113.16      0.95   119.35   0.00
-------------------------------------------------------

# Marginal effects data frames
ggpredict_1 <- ggpredict(lm_pre_alpha, terms = "area_living") # total model
ggpredict_2 <- ggpredict(lm_pre_alpha_area, terms = "area_living") # non-linear addition
ggpredict_3 <- ggpredict(lm_pre_alpha_area_single, terms = "area_living") # single-variable fit

# Plots
b <- ggplot(data_factor_core, aes(x = area_living)) +
   geom_smooth(data_factor, mapping = aes(y = sold_price), color = "grey50") +
   geom_smooth(ggpredict_1, mapping = aes(x, predicted), linetype = "dashed", color = very_low) +
   geom_smooth(ggpredict_2, mapping = aes(x, predicted), linetype = "dashed", color = med) +
     labs(title = "Living Area and Price",
          x = "Living Area",
          y = "Prediction")

# Look at age & age^2 alone to see impact on more relevant y-axis scale
c <- ggplot() +
   geom_smooth(ggpredict_1, mapping = aes(x, predicted), linetype = "dashed", color = very_low) +
   geom_smooth(ggpredict_2, mapping = aes(x, predicted), linetype = "dashed", color = med) +
     labs(title = "Living Area and Price",
          x = "Living Area",
          y = "Prediction")

# Conclusion
a

`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'
`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'

gridExtra::grid.arrange(b,c, nrow =2, ncol = 1)

`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'
`geom_smooth()` using method = 'loess' and formula 'y ~ x'
`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'
`geom_smooth()` using method = 'loess' and formula 'y ~ x'
`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'

1.3.3 Land

# General graphing
ggplot(data_factor, aes(x = land_acres , y = sold_price)) + 
    geom_point(aes(color = infections_period), alpha = 0.15) + 
    geom_smooth(aes(color = infections_period)) +
    geom_smooth(color = "grey50", linetype = "dashed") +
    theme_minimal()

ggplot(data_factor, aes(x = land_acres, y = sold_price/land_acres)) + 
    geom_point(aes(color = infections_period), alpha = 0.15) + 
    geom_smooth(aes(color = infections_period)) +
    geom_smooth(color = "grey50", linetype = "dashed") +
    theme_minimal()

1.3.4 Non-linear Additions

#Additions
data_factor_core_clean <- data_factor_core
data_factor_core_clean$age_2 <- I(data_factor_core$age^2)
data_factor_core_clean$area_living_2 <- I(data_factor_core$area_living^2)

1.4 Accounting for Multicollinearity

# Full model summary
summ(lm_pre_alpha)

# Check Variance Inflation Factors (VIF)
VIF(lm_pre_alpha)
alias(lm_pre_alpha)

# Total area and living area are found to be significantly (i.e. VIF > 5) multicolinear (expected)
# Solution: Remove area_total

# Note the significant drop in R^2 from 0.99 to 0.86
lm_pre_alpha_cleaned <- lm(log(sold_price) ~ . - area_total ,data = data_factor_core)
summ(lm_pre_alpha_cleaned)
VIF(lm_pre_alpha_cleaned)

# Final pre_alpha
VIF(lm_pre_alpha_cleaned)
alias(lm_pre_alpha_cleaned)

# Another way to check for multicollinearity is visually through the mcvis package
data_numeric <- select_if(data_factor_core, is.numeric) # Subset numeric columns with dplyr
mcvis_result <- mcvis(X = data_numeric)
a <- plot(mcvis_result)

par(mfrow = c(2,2))
#Removals
data_numeric <- subset(data_numeric, select = -c(list_price))
mcvis_result <- mcvis(X = data_numeric)
b <- plot(mcvis_result)

#Removals
data_numeric <- subset(data_numeric, select = -c(area_total))
mcvis_result <- mcvis(X = data_numeric)
c <- plot(mcvis_result)

a
b
c

1.4.1 Multicollinearity Removals

# Removals
# - Area_total
# - Listing price

par(mfrow = c(2,2))

data_factor_core_clean <- subset(data_factor_core_clean, select = -c(area_total, list_price))

1.5 High-leverage Removals

data_factor_core_clean <- data_factor_core_clean[-c(23515), ]

1.5 Alpha Model

cl <- makePSOCKcluster(5)
registerDoParallel(cl)
tab_model(lm_alpha, ci_method = "wald")

Profiled confidence intervals may take longer time to compute. Use 'ci_method="wald"' for faster computation of CIs.

2. Factor Analysis

2.1 Corona

2.1.1 Visualization

# Waves of infection
ggplot(data_factor, aes(x = as.Date(sold_date), y = infections_3mma)) + 
    geom_point(color = low, alpha = 0.7) + 
    geom_smooth(linetype = "dashed", color = med) +
    theme_minimal() +
    scale_x_date(limits = as.Date(c("2020-01-01", "2021-12-31"))) +
    scale_y_continuous(limits = c(0,max(infections_3mma))) +
    xlab(" ") +
    ylab("Confirmed Infections per Day") +
    labs(title = "Waves of Infection",
         caption = "") +
    geom_vline(xintercept = as.numeric(as.Date("2020-03-23")), linetype=4)

`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'
Warning: Removed 17731 rows containing non-finite values (stat_smooth).
Warning: Removed 17731 rows containing missing values (geom_point).
Warning: Removed 3 rows containing missing values (geom_smooth).

# Accumulation of infections
ggplot(data_factor, aes(x = as.Date(sold_date), y = I(infections_accum/1000))) + 
    geom_point(color = low, alpha = 0.7) + 
    geom_smooth(linetype = "dashed", color = med) +
    theme_minimal() +
    scale_x_date(limits = as.Date(c("2020-01-01", "2021-12-31"))) +
    scale_y_continuous(limits = c(0,max(I(infections_accum/1000)))) +
    xlab(" ") +
    ylab("Accumulation of Infections (in 000's)") +
    labs(title = "Accumulation of Infections",
         caption = "")

`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'
Warning: Removed 17731 rows containing non-finite values (stat_smooth).
Warning: Removed 17731 rows containing missing values (geom_point).
Warning: Removed 3 rows containing missing values (geom_smooth).

# Infections and home prices
ggplot(data_factor, aes(x = I(infections_3mma/1000), y = sold_price)) + 
    #geom_point() + 
    geom_smooth(linetype = "dashed", color = med) +
    theme_minimal() +
    scale_x_continuous( limits = c(0,max(I(infections_3mma/1000)))) +
    xlab("3-Month Moving Average of Daily Infections (in 000's)") +
    ylab("Sold Price (Actual)") +
    labs(title = "Infections and Price",
         caption = "")

`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'

# "#ff6c67", "#00c2c6"

ggplot(data_factor, aes(x = infections_period, y = sold_price/1000, fill = infections_period)) +
    geom_violin(alpha = 0.5) +
    geom_boxplot(width=0.1) +
    scale_fill_viridis(discrete = TRUE, alpha=0.6, option="D") +
    coord_flip() +
    theme_ipsum() +
    theme(
      legend.position="none") +
    xlab("Infections Present (1 = yes)") +
    ylab("Sold Price (in 000's)") +
    scale_fill_manual(values=c(very_low, med)) +
    labs(title = "Comparison of Sold Price",
         caption = "e")

Scale for 'fill' is already present. Adding another scale for 'fill', which will replace the existing scale.

2.1.2 Modeling

# Plots
ggplot(data_factor_core, aes(x = infections_3mma)) +
   geom_smooth(data_factor_core, mapping = aes(y = sold_price), color = "grey50") + # Actual Data
   geom_smooth(ggpredict_1, mapping = aes(x, predicted), linetype = "dashed", color = low) + # Controlled model
   geom_smooth(ggpredict_2, mapping = aes(x, predicted), linetype = "dashed", color = med) + # Best single fit
   ggtitle("Model Fit Overview")

`geom_smooth()` using method = 'gam' and formula 'y ~ s(x, bs = "cs")'
`geom_smooth()` using method = 'loess' and formula 'y ~ x'
`geom_smooth()` using method = 'loess' and formula 'y ~ x'

2.2 Corona on Number of Bedrooms

2.2.1 Visualiztion

gridExtra::grid.arrange(a)
gridExtra::grid.arrange(b)

gridExtra::grid.arrange(c)

gridExtra::grid.arrange(d)

gridExtra::grid.arrange(e)

2.2.2 Modeling

Ideas

• Break into each room number

coeftest(lm_corona_bedrooms, vcov = vcovHC(lm_corona_bedrooms, method = "White2", type = "HC0"))

t test of coefficients:

                                           Estimate  Std. Error  t value  Pr(>|t|)    
(Intercept)                              2.3414e+05  3.0157e+04   7.7640 8.553e-15 ***
ac_typenone                             -5.4498e+04  1.9525e+03 -27.9111 < 2.2e-16 ***
ac_typenot_central                      -2.2087e+04  1.6931e+03 -13.0457 < 2.2e-16 ***
patio1                                   1.2472e+04  8.3412e+02  14.9527 < 2.2e-16 ***
school_general1                          8.2759e+03  1.0945e+03   7.5614 4.129e-14 ***
photo_count                              1.3248e+03  4.9480e+01  26.7740 < 2.2e-16 ***
pool1                                    1.9075e+04  1.5001e+03  12.7159 < 2.2e-16 ***
roof_typeother                           7.3681e+03  1.4621e+03   5.0392 4.708e-07 ***
roof_typeshingle                         2.7197e+04  1.6956e+03  16.0396 < 2.2e-16 ***
roof_typeslate                           1.5496e+04  9.0682e+03   1.7088 0.0874995 .  
gas_typenatural                         -1.0756e+05  3.4559e+03 -31.1233 < 2.2e-16 ***
gas_typenone                            -1.3865e+05  2.2971e+03 -60.3594 < 2.2e-16 ***
gas_typepropane                         -9.3236e+04  1.8180e+04  -5.1285 2.943e-07 ***
gas_typeunknown                         -1.3842e+05  2.1427e+03 -64.5988 < 2.2e-16 ***
out_building1                           -5.5192e+03  8.8805e+02  -6.2149 5.218e-10 ***
appliances1                              2.5898e+04  1.1928e+03  21.7118 < 2.2e-16 ***
property_conditionnew                   -2.0935e+04  6.3471e+03  -3.2983 0.0009741 ***
property_conditionother                 -2.0956e+04  1.0429e+03 -20.0948 < 2.2e-16 ***
energy_efficient1                        1.8928e+04  8.8970e+02  21.2746 < 2.2e-16 ***
exterior_typemetal                      -4.0964e+03  2.4309e+03  -1.6852 0.0919667 .  
exterior_typeother                       1.3327e+04  1.1559e+03  11.5302 < 2.2e-16 ***
exterior_typevinyl                       3.0630e+03  1.2148e+03   2.5213 0.0116992 *  
exterior_typewood                        6.8287e+02  1.8873e+03   0.3618 0.7174878    
exterior_featurescourtyard               3.8981e+04  1.4928e+04   2.6113 0.0090249 ** 
exterior_featuresfence                  -2.3394e+04  5.4658e+03  -4.2800 1.876e-05 ***
exterior_featuresnone                   -1.3995e+04  5.4825e+03  -2.5528 0.0106928 *  
exterior_featuresporch                  -2.0091e+04  5.5495e+03  -3.6203 0.0002948 ***
exterior_featurestennis_court            2.3977e+04  1.3892e+04   1.7260 0.0843658 .  
fireplace1                               3.1903e+04  8.3534e+02  38.1915 < 2.2e-16 ***
foundation_typeslab                      2.0170e+04  1.3210e+03  15.2687 < 2.2e-16 ***
foundation_typeunspecified               9.7919e+03  1.4755e+03   6.6362 3.286e-11 ***
beds_total1                             -7.1968e+04  2.9707e+04  -2.4226 0.0154172 *  
beds_total2                             -5.4848e+04  2.9462e+04  -1.8616 0.0626666 .  
beds_total3                             -2.8699e+04  2.9463e+04  -0.9741 0.3300308    
beds_total4                              1.0509e+04  2.9483e+04   0.3564 0.7215164    
beds_total5                              1.7689e+04  3.0042e+04   0.5888 0.5559999    
age                                     -2.1730e+03  8.0380e+01 -27.0343 < 2.2e-16 ***
dom                                      8.3326e+00  6.9823e+00   1.1934 0.2327235    
sewer_typeseptic                        -4.5359e+03  1.5203e+03  -2.9836 0.0028514 ** 
sewer_typeunspecified                   -4.4686e+03  8.1555e+02  -5.4792 4.314e-08 ***
property_stylenot_mobile                 7.2464e+04  1.8129e+03  39.9708 < 2.2e-16 ***
subdivision1                             2.7805e+03  9.7726e+02   2.8452 0.0044416 ** 
water_typewell                          -3.2549e+03  4.4298e+03  -0.7348 0.4624852    
waterfront1                              2.7540e+04  1.6343e+03  16.8507 < 2.2e-16 ***
bottom25_dom1                            1.2415e+04  1.0822e+03  11.4716 < 2.2e-16 ***
age_2                                    1.9417e+01  1.1018e+00  17.6236 < 2.2e-16 ***
data_factor$infections_3mma             -2.8687e+01  1.5003e+01  -1.9121 0.0558708 .  
beds_total1:data_factor$infections_3mma  2.5212e+01  1.5454e+01   1.6314 0.1028173    
beds_total2:data_factor$infections_3mma  3.2225e+01  1.5051e+01   2.1411 0.0322755 *  
beds_total3:data_factor$infections_3mma  3.6766e+01  1.5014e+01   2.4488 0.0143384 *  
beds_total4:data_factor$infections_3mma  3.7085e+01  1.5053e+01   2.4637 0.0137574 *  
beds_total5:data_factor$infections_3mma  4.7405e+01  1.6056e+01   2.9525 0.0031552 ** 
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

2.3 Corona on Price Quantiles

2.3.1 Visualization

# Find the mean of each group
library(plyr)
price_means <- ddply(data_factor, "infections_period", summarise, price_mean = mean(sold_price, na.rm = TRUE))

# Distribution: Total
ggplot(data_factor, aes(x = sold_price)) +
    geom_density(alpha = 0.5, position = "identity", fill = very_low) +
    ggtitle("Price Distribution") +
    geom_vline(data=price_means, aes(xintercept = mean(sold_price)), linetype="dashed", size= 0.4, color = very_low, alpha = 0.8) +
    xlab("Sold Price") +
    ylab("Density") 

# Distribution: Infection
ggplot(data_factor, aes(x = sold_price, fill = infections_period)) +
    geom_density(alpha = 0.5, position = "identity") +
    ggtitle("Price Distributions") +
    geom_vline(data=price_means, aes(xintercept = price_means[2,2]), linetype="dashed", size= 0.4, color = med, alpha = 0.8) +
    geom_vline(data = price_means, aes(xintercept = price_means[1,2]), linetype="dashed", size= 0.4, color = very_low, alpha = 0.8) +
    scale_fill_manual(values = c(very_low, med), 
                      name = "Infection Period",
                      labels = c("Pre", "Post")) +
    xlab("Sold Price") +
    ylab("Density") +
    labs(fill = "Infection Period")

# Distribution: Top vs. Bottom
ggplot(data_factor) +
    geom_density(aes(x = sold_price, fill = infections_period), alpha = 0.5, position = "identity") + 
    facet_grid(vars(top25_sold_price, bottom25_sold_price), scales = "free") +
    ggtitle("Price Distributions") +
    scale_fill_manual(values=c(very_low, med)) +
    xlab("Sold Price") +
    labs(fill = "Infection Period") +
    ylab("Density") +
    scale_fill_manual(values = c(very_low, med), 
                      name = "Infection Period",
                      labels = c("Pre", "Post"))

#Price and Infections
ggplot(data_factor, aes(x = infections_period, y = sold_price, fill = infections_period)) +
    geom_violin(alpha = 0.5) +
    geom_boxplot(width=0.1) +
    scale_fill_viridis(discrete = TRUE, alpha=0.6, option="D") +
    coord_flip() +
    theme_ipsum() +
    theme(
      legend.position="none",
      plot.title = element_text(size=11)) +
    ggtitle("Comparison of Sold Price") +
    xlab("Infection Period") +
    scale_fill_manual(values=c(very_low, med)) +
    ylab("Sold Price") 

2.3.2 Modeling

coeftest(lm_corona_price_bottom, vcov = vcovHC(lm_corona_price_bottom, method = "White2", type = "HC0"))

t test of coefficients:

                                                   Estimate  Std. Error  t value  Pr(>|t|)    
(Intercept)                                      2.6017e+05  2.1950e+04  11.8531 < 2.2e-16 ***
property_typeDUP                                -2.1684e+04  1.5731e+04  -1.3784 0.1680918    
property_typeOTH                                 7.3064e+03  8.1447e+03   0.8971 0.3696882    
property_typePAT                                 9.6853e+03  4.8857e+03   1.9824 0.0474454 *  
property_typeSGL                                 1.8003e+04  2.2688e+03   7.9351 2.194e-15 ***
property_typeTNH                                -4.5428e+03  2.8170e+03  -1.6126 0.1068353    
ac_typenone                                     -2.4963e+04  1.3345e+03 -18.7056 < 2.2e-16 ***
ac_typenot_central                              -3.5920e+03  1.2040e+03  -2.9833 0.0028541 ** 
patio1                                           4.1811e+03  6.4168e+02   6.5158 7.372e-11 ***
school_general1                                  6.8812e+03  8.4726e+02   8.1217 4.811e-16 ***
photo_count                                      5.9194e+02  3.8741e+01  15.2795 < 2.2e-16 ***
pool1                                            1.0669e+04  1.2351e+03   8.6380 < 2.2e-16 ***
roof_typeother                                  -3.3318e+02  1.1443e+03  -0.2912 0.7709303    
roof_typeshingle                                 1.1199e+04  1.3489e+03   8.3025 < 2.2e-16 ***
roof_typeslate                                   2.6920e+03  7.2137e+03   0.3732 0.7090169    
gas_typenatural                                 -7.3831e+04  2.9774e+03 -24.7972 < 2.2e-16 ***
gas_typenone                                    -1.0662e+05  2.0419e+03 -52.2141 < 2.2e-16 ***
gas_typepropane                                 -6.9963e+04  1.4886e+04  -4.7000 2.616e-06 ***
gas_typeunknown                                 -1.0796e+05  1.9714e+03 -54.7669 < 2.2e-16 ***
out_building1                                   -6.4916e+03  6.8667e+02  -9.4537 < 2.2e-16 ***
area_living                                     -6.9603e+00  5.2515e+00  -1.3254 0.1850514    
land_acres                                       1.9796e+03  7.4583e+02   2.6542 0.0079554 ** 
appliances1                                      1.0788e+04  8.4929e+02  12.7019 < 2.2e-16 ***
garage1                                          6.8509e+03  6.3401e+02  10.8056 < 2.2e-16 ***
property_conditionnew                           -8.4140e+03  5.3106e+03  -1.5844 0.1131184    
property_conditionother                         -1.0277e+04  8.3230e+02 -12.3471 < 2.2e-16 ***
energy_efficient1                                1.0499e+04  7.0947e+02  14.7989 < 2.2e-16 ***
exterior_typemetal                              -8.7770e+02  1.8545e+03  -0.4733 0.6360072    
exterior_typeother                               7.9162e+03  8.8607e+02   8.9341 < 2.2e-16 ***
exterior_typevinyl                               1.8030e+03  9.2587e+02   1.9474 0.0515005 .  
exterior_typewood                                2.4987e+03  1.3755e+03   1.8166 0.0692898 .  
exterior_featurescourtyard                       2.3714e+04  1.2638e+04   1.8764 0.0606085 .  
exterior_featuresfence                          -2.4910e+04  4.2901e+03  -5.8063 6.466e-09 ***
exterior_featuresnone                           -2.0388e+04  4.2886e+03  -4.7539 2.007e-06 ***
exterior_featuresporch                          -2.4831e+04  4.3365e+03  -5.7260 1.040e-08 ***
exterior_featurestennis_court                    2.3169e+03  1.0105e+04   0.2293 0.8186463    
fireplace1                                       1.0708e+04  6.8071e+02  15.7301 < 2.2e-16 ***
foundation_typeslab                              4.5790e+03  1.0141e+03   4.5155 6.348e-06 ***
foundation_typeunspecified                       2.5181e+03  1.0937e+03   2.3023 0.0213285 *  
beds_total1                                     -8.8157e+03  2.0732e+04  -0.4252 0.6706851    
beds_total2                                     -1.8691e+04  2.0643e+04  -0.9055 0.3652225    
beds_total3                                     -2.6547e+04  2.0670e+04  -1.2843 0.1990422    
beds_total4                                     -2.0696e+04  2.0696e+04  -1.0000 0.3173106    
beds_total5                                     -3.5958e+04  2.1091e+04  -1.7049 0.0882328 .  
bath_full1                                      -1.4452e+04  1.3423e+04  -1.0766 0.2816580    
bath_full2                                      -7.5442e+03  1.3405e+04  -0.5628 0.5735714    
bath_full3                                       1.4197e+04  1.3524e+04   1.0497 0.2938466    
bath_full4                                       7.3672e+03  2.0346e+04   0.3621 0.7172836    
bath_full6                                       4.0238e+04  1.4399e+04   2.7945 0.0052018 ** 
bath_half1                                       1.1964e+04  9.6620e+02  12.3824 < 2.2e-16 ***
bath_half2                                       2.3815e+04  7.0879e+03   3.3599 0.0007808 ***
bath_half3                                       5.8803e+04  9.6367e+03   6.1020 1.063e-09 ***
bath_half4                                       1.0372e+05  2.8037e+03  36.9928 < 2.2e-16 ***
bath_half5                                      -2.7354e+04  2.1637e+04  -1.2642 0.2061652    
age                                             -1.5156e+03  6.6548e+01 -22.7751 < 2.2e-16 ***
dom                                             -1.1463e+01  5.3153e+00  -2.1567 0.0310414 *  
sewer_typeseptic                                -6.1909e+03  1.1518e+03  -5.3749 7.732e-08 ***
sewer_typeunspecified                           -4.5154e+03  6.2787e+02  -7.1916 6.591e-13 ***
property_stylenot_mobile                         2.8815e+04  1.5438e+03  18.6650 < 2.2e-16 ***
subdivision1                                     2.4505e+03  7.3353e+02   3.3407 0.0008370 ***
water_typewell                                   2.4004e+03  3.2270e+03   0.7438 0.4569754    
waterfront1                                      1.7312e+04  1.2796e+03  13.5289 < 2.2e-16 ***
bottom25_dom1                                    7.6382e+03  8.1707e+02   9.3483 < 2.2e-16 ***
age_2                                            1.3323e+01  9.1771e-01  14.5177 < 2.2e-16 ***
area_living_2                                    1.5910e-02  1.5353e-03  10.3627 < 2.2e-16 ***
data_factor$infections_3mma                      7.8359e+00  4.6475e-01  16.8604 < 2.2e-16 ***
bottom25_sold_price                             -7.9381e+04  7.9732e+02 -99.5596 < 2.2e-16 ***
data_factor$infections_3mma:bottom25_sold_price -5.1963e+00  7.5026e-01  -6.9259 4.440e-12 ***
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

2.4 Corona on Age Quantiles

2.4.1 Visualization

#Age on Infections
ggplot(data_factor, aes(x = infections_period, y = age, fill = infections_period)) +
    geom_violin(alpha = 0.5) +
    geom_boxplot(width=0.1) +
    scale_fill_viridis(discrete = TRUE, alpha=0.6, option="D") +
    coord_flip() +
    theme_ipsum() +
    theme(
      legend.position="none",
      plot.title = element_text(size=14)) +
    ggtitle("Comparison of Age") +
    xlab("Infection Period") +
    ylab("Age of Property") +
    scale_fill_manual(values = c(very_low, med), 
                      name = "Infection Period",
                      labels = c("Pre", "Post"))

Scale for 'fill' is already present. Adding another scale for 'fill', which will replace the existing scale.

2.4.2 Modeling

coeftest(lm_corona_age_bottom, vcov = vcovHC(lm_corona_age_bottom, method = "White2", type = "HC0"))

t test of coefficients:

                                            Estimate  Std. Error  t value  Pr(>|t|)    
(Intercept)                               1.2697e+05  3.1235e+04   4.0649 4.821e-05 ***
ac_typenone                              -4.4551e+04  1.9540e+03 -22.8002 < 2.2e-16 ***
ac_typenot_central                       -1.4115e+04  1.5258e+03  -9.2508 < 2.2e-16 ***
patio1                                    8.8898e+03  7.5883e+02  11.7151 < 2.2e-16 ***
school_general1                           1.0875e+04  1.0179e+03  10.6833 < 2.2e-16 ***
photo_count                               8.7047e+02  4.7073e+01  18.4919 < 2.2e-16 ***
pool1                                     8.6980e+03  1.3425e+03   6.4787 9.426e-11 ***
roof_typeother                            3.2938e+03  1.4117e+03   2.3331 0.0196485 *  
roof_typeshingle                          2.1981e+04  1.6190e+03  13.5770 < 2.2e-16 ***
roof_typeslate                            7.4667e+03  8.7967e+03   0.8488 0.3959985    
gas_typenatural                          -9.3517e+04  3.5479e+03 -26.3584 < 2.2e-16 ***
gas_typenone                             -1.2593e+05  2.5036e+03 -50.3013 < 2.2e-16 ***
gas_typepropane                          -9.3374e+04  1.8322e+04  -5.0962 3.491e-07 ***
gas_typeunknown                          -1.2906e+05  2.4005e+03 -53.7657 < 2.2e-16 ***
out_building1                            -6.3411e+03  8.0363e+02  -7.8906 3.132e-15 ***
land_acres                                3.7274e+03  9.3788e+02   3.9743 7.080e-05 ***
appliances1                               2.5344e+04  1.1141e+03  22.7473 < 2.2e-16 ***
garage1                                   1.3975e+04  7.4840e+02  18.6729 < 2.2e-16 ***
property_conditionnew                    -5.0567e+03  6.2096e+03  -0.8143 0.4154609    
property_conditionother                  -2.0425e+04  9.3375e+02 -21.8744 < 2.2e-16 ***
energy_efficient1                         1.5099e+04  8.2260e+02  18.3556 < 2.2e-16 ***
exterior_typemetal                       -2.5826e+02  2.3574e+03  -0.1096 0.9127626    
exterior_typeother                        1.2023e+04  1.0621e+03  11.3193 < 2.2e-16 ***
exterior_typevinyl                        5.5811e+03  1.0999e+03   5.0740 3.924e-07 ***
exterior_typewood                         3.2842e+03  1.7401e+03   1.8874 0.0591170 .  
exterior_featurescourtyard                4.5132e+04  1.5039e+04   3.0009 0.0026945 ** 
exterior_featuresfence                   -1.4935e+04  4.9711e+03  -3.0044 0.0026641 ** 
exterior_featuresnone                    -7.0656e+03  4.9869e+03  -1.4168 0.1565455    
exterior_featuresporch                   -1.2793e+04  5.0376e+03  -2.5394 0.0111099 *  
exterior_featurestennis_court             1.9681e+04  1.0734e+04   1.8335 0.0667358 .  
fireplace1                                1.2147e+04  8.0986e+02  14.9993 < 2.2e-16 ***
foundation_typeslab                       1.3287e+04  1.2600e+03  10.5449 < 2.2e-16 ***
foundation_typeunspecified                7.2193e+03  1.4077e+03   5.1285 2.943e-07 ***
beds_total1                              -2.4876e+04  2.7454e+04  -0.9061 0.3649013    
beds_total2                              -2.7848e+04  2.7266e+04  -1.0214 0.3070895    
beds_total3                              -2.5167e+04  2.7263e+04  -0.9231 0.3559582    
beds_total4                              -2.0523e+04  2.7292e+04  -0.7520 0.4520777    
beds_total5                              -3.6440e+04  2.7717e+04  -1.3147 0.1886165    
bath_full1                               -3.8425e+04  2.3952e+04  -1.6043 0.1086671    
bath_full2                               -1.3304e+04  2.3946e+04  -0.5556 0.5784882    
bath_full3                                6.7324e+03  2.4019e+04   0.2803 0.7792514    
bath_full4                                1.1122e+03  2.9704e+04   0.0374 0.9701332    
bath_full6                               -9.3526e+03  2.4547e+04  -0.3810 0.7031968    
bath_half1                                1.0957e+04  1.0886e+03  10.0654 < 2.2e-16 ***
bath_half2                                3.0347e+04  6.6151e+03   4.5875 4.508e-06 ***
bath_half3                                6.3102e+04  9.8354e+03   6.4158 1.427e-10 ***
bath_half4                                8.8086e+04  3.1342e+03  28.1044 < 2.2e-16 ***
bath_half5                               -5.2038e+04  2.5797e+04  -2.0172 0.0436870 *  
dom                                      -2.3526e+01  6.3618e+00  -3.6980 0.0002178 ***
sold_date                                 1.2018e+00  4.5403e-01   2.6470 0.0081272 ** 
sewer_typeseptic                         -6.1146e+03  1.4223e+03  -4.2992 1.721e-05 ***
sewer_typeunspecified                    -3.7371e+03  7.3964e+02  -5.0526 4.389e-07 ***
property_stylenot_mobile                  6.9362e+04  1.7331e+03  40.0209 < 2.2e-16 ***
subdivision1                              3.5289e+03  9.0359e+02   3.9054 9.432e-05 ***
water_typewell                           -5.1055e+02  3.9561e+03  -0.1291 0.8973140    
waterfront1                               2.0044e+04  1.4711e+03  13.6256 < 2.2e-16 ***
bottom25_dom1                             1.0831e+04  9.7524e+02  11.1063 < 2.2e-16 ***
area_living_2                             1.6847e-02  4.0061e-04  42.0532 < 2.2e-16 ***
data_factor$infections_3mma               8.6802e+00  6.9100e-01  12.5617 < 2.2e-16 ***
bottom25_age                              2.5461e+04  9.3603e+02  27.2015 < 2.2e-16 ***
data_factor$infections_3mma:bottom25_age  7.9891e-01  8.6329e-01   0.9254 0.3547531    
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

2.5 Corona on Size Quantiles

2.5.1 Visualization

#area_living on Infections
ggplot(data_factor, aes(x = infections_period, y = sold_price/area_living, fill = infections_period)) +
    geom_violin(alpha = 0.5) +
    geom_boxplot(width=0.1) +
    scale_fill_viridis(discrete = TRUE, alpha=0.6, option="D") +
    coord_flip() +
    theme_ipsum() +
    theme(
      legend.position="none",
      plot.title = element_text(size=11)) +
    ggtitle("Comparison of Living Area per Sqft.") +
    xlab("Infection Period") +
    ylab("Price per Living Area") +
    scale_fill_manual(values=c(very_low, med)) +
    scale_y_continuous(limits = c(0,250))

Scale for 'fill' is already present. Adding another scale for 'fill', which will replace the existing scale.
Warning: Removed 68 rows containing non-finite values (stat_ydensity).
Warning: Removed 68 rows containing non-finite values (stat_boxplot).

2.5.2 Modeling

coeftest(lm_corona_area_living_top, vcov = vcovHC(lm_corona_area_living_top, method = "White2", type = "HC0"))

t test of coefficients:

                                                 Estimate  Std. Error  t value  Pr(>|t|)    
(Intercept)                                    2.1786e+05  3.1676e+04   6.8776 6.232e-12 ***
ac_typenone                                   -4.5055e+04  1.9559e+03 -23.0347 < 2.2e-16 ***
ac_typenot_central                            -1.4868e+04  1.5495e+03  -9.5953 < 2.2e-16 ***
patio1                                         9.0080e+03  7.7344e+02  11.6466 < 2.2e-16 ***
school_general1                                8.5012e+03  1.0371e+03   8.1972 2.580e-16 ***
photo_count                                    1.0514e+03  4.8373e+01  21.7341 < 2.2e-16 ***
pool1                                          1.1293e+04  1.3644e+03   8.2768 < 2.2e-16 ***
roof_typeother                                 4.4255e+03  1.4022e+03   3.1562  0.001601 ** 
roof_typeshingle                               2.2060e+04  1.6174e+03  13.6392 < 2.2e-16 ***
roof_typeslate                                 7.5155e+03  8.9239e+03   0.8422  0.399699    
gas_typenatural                               -1.0170e+05  3.5575e+03 -28.5866 < 2.2e-16 ***
gas_typenone                                  -1.3662e+05  2.4951e+03 -54.7537 < 2.2e-16 ***
gas_typepropane                               -1.0682e+05  1.7037e+04  -6.2695 3.682e-10 ***
gas_typeunknown                               -1.3645e+05  2.3853e+03 -57.2035 < 2.2e-16 ***
out_building1                                 -4.6203e+03  8.2442e+02  -5.6043 2.114e-08 ***
land_acres                                     5.3460e+03  9.4296e+02   5.6694 1.449e-08 ***
appliances1                                    2.4211e+04  1.1233e+03  21.5532 < 2.2e-16 ***
garage1                                        1.4565e+04  7.6157e+02  19.1246 < 2.2e-16 ***
property_conditionnew                         -1.8401e+04  6.3472e+03  -2.8992  0.003745 ** 
property_conditionother                       -1.9911e+04  9.6101e+02 -20.7193 < 2.2e-16 ***
energy_efficient1                              1.4472e+04  8.3430e+02  17.3469 < 2.2e-16 ***
exterior_typemetal                            -1.9967e+03  2.3583e+03  -0.8466  0.397202    
exterior_typeother                             1.1470e+04  1.0702e+03  10.7168 < 2.2e-16 ***
exterior_typevinyl                             3.3346e+03  1.1187e+03   2.9807  0.002878 ** 
exterior_typewood                              1.4095e+03  1.7583e+03   0.8016  0.422795    
exterior_featurescourtyard                     4.3545e+04  1.4388e+04   3.0266  0.002476 ** 
exterior_featuresfence                        -1.3817e+04  5.1863e+03  -2.6641  0.007724 ** 
exterior_featuresnone                         -6.5726e+03  5.2004e+03  -1.2639  0.206292    
exterior_featuresporch                        -1.3384e+04  5.2569e+03  -2.5460  0.010902 *  
exterior_featurestennis_court                  2.0086e+04  1.1798e+04   1.7025  0.088672 .  
fireplace1                                     1.9245e+04  7.9863e+02  24.0978 < 2.2e-16 ***
foundation_typeslab                            1.3112e+04  1.2770e+03  10.2682 < 2.2e-16 ***
foundation_typeunspecified                     7.1261e+03  1.4117e+03   5.0479 4.499e-07 ***
beds_total1                                   -2.4323e+04  2.8620e+04  -0.8499  0.395408    
beds_total2                                   -2.0505e+04  2.8451e+04  -0.7207  0.471080    
beds_total3                                   -8.6578e+03  2.8444e+04  -0.3044  0.760845    
beds_total4                                    2.7001e+03  2.8461e+04   0.0949  0.924420    
beds_total5                                   -4.1571e+03  2.8867e+04  -0.1440  0.885494    
bath_full1                                    -6.0233e+04  2.5477e+04  -2.3642  0.018076 *  
bath_full2                                    -2.2302e+04  2.5477e+04  -0.8754  0.381375    
bath_full3                                     7.4722e+03  2.5549e+04   0.2925  0.769934    
bath_full4                                     1.4402e+04  3.1333e+04   0.4596  0.645772    
bath_full6                                    -2.8661e+04  2.6137e+04  -1.0965  0.272850    
bath_half1                                     1.6583e+04  1.0888e+03  15.2302 < 2.2e-16 ***
bath_half2                                     4.0757e+04  6.9462e+03   5.8675 4.482e-09 ***
bath_half3                                     7.1768e+04  1.0544e+04   6.8065 1.023e-11 ***
bath_half4                                     7.0281e+04  3.5280e+03  19.9209 < 2.2e-16 ***
bath_half5                                    -4.1415e+04  4.2895e+04  -0.9655  0.334309    
age                                           -1.9822e+03  7.9428e+01 -24.9554 < 2.2e-16 ***
dom                                           -1.5428e+01  6.4584e+00  -2.3888  0.016913 *  
sold_date                                      6.1595e-01  4.7569e-01   1.2949  0.195377    
sewer_typeseptic                              -6.0698e+03  1.4378e+03  -4.2215 2.436e-05 ***
sewer_typeunspecified                         -4.9915e+03  7.5983e+02  -6.5692 5.161e-11 ***
property_stylenot_mobile                       7.3425e+04  1.7430e+03  42.1247 < 2.2e-16 ***
subdivision1                                   2.8893e+03  9.1343e+02   3.1631  0.001563 ** 
water_typewell                                 1.6295e+03  4.0353e+03   0.4038  0.686356    
waterfront1                                    2.0708e+04  1.4983e+03  13.8212 < 2.2e-16 ***
bottom25_dom1                                  1.1119e+04  9.9811e+02  11.1405 < 2.2e-16 ***
age_2                                          1.8417e+01  1.0995e+00  16.7510 < 2.2e-16 ***
data_factor$infections_3mma                    8.6770e+00  5.9698e-01  14.5349 < 2.2e-16 ***
top25_area_living                              3.8537e+04  1.3446e+03  28.6604 < 2.2e-16 ***
data_factor$infections_3mma:top25_area_living  9.7205e-01  1.2004e+00   0.8098  0.418087    
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

2.6 Corona on Days on Market

2.6.1 Visualization

# Conditional Mean
library(plyr)
dom_mean_data <- ddply(data_factor, "infections_period", summarise, dom_mean = mean(dom, na.rm = TRUE))

# Distribution: Just for City
ggplot(data_factor, aes(x = dom)) +
    geom_density(alpha = 0.5, position = "identity", fill = very_low) +
    ggtitle("Days on Market Distribution") +
    geom_vline(aes(xintercept = mean(dom)), linetype="dashed", size= 0.4, alpha = 0.5, color = very_low) +
    xlab("Days on Market") +
    ylab("Density")

# Distribution: Infection
ggplot(data_factor, aes(x = dom, fill = infections_period)) +
    geom_density(alpha = 0.5, position = "identity") +
    ggtitle("Days on Market Distributions") +
    geom_vline(data = dom_mean_data, aes(xintercept = dom_mean_data[2,2]), linetype="dashed", size= 0.5, color = med, alpha = 0.8) +
    geom_vline(data = dom_mean_data, aes(xintercept = dom_mean_data[1,2]), linetype="dashed", size= 0.5, alpha = 0.8, color = very_low) +
    xlab("Days on Market") +
    ylab("Density") +
    scale_fill_manual(values = c(very_low, med), 
                      name = "Infection Period",
                      labels = c("Pre", "Post"))

# Distribution: Top vs. Bottom
ggplot(data_factor) +
    geom_density(aes(x = dom, fill = infections_period), alpha = 0.5, position = "identity") + 
    facet_grid(vars(top25_dom, bottom25_dom), scales = "free") +
    ggtitle("Days on Market Distributions") +
    xlab("Days on Market") +
    ylab("Density") +
    scale_fill_manual(values = c(very_low, med), 
                      name = "Infection Period",
                      labels = c("Pre", "Post"))

#dom on Infections
ggplot(data_factor, aes(x = infections_period, y = dom, fill = infections_period)) +
    geom_violin(alpha = 0.5) +
    geom_boxplot(width=0.1) +
    scale_fill_viridis(discrete = TRUE, alpha=0.6, option="D") +
    #coord_flip() +
    theme_ipsum() +
    theme(
      legend.position="none",
      plot.title = element_text(size=11)) +
    ggtitle("Comparison of Days on Market") +
    xlab("Infection Period") +
    ylab("Days on Market") +
    scale_fill_manual(values = c(very_low, med), 
                      name = "Infection Period",
                      labels = c("Pre", "Post"))

Scale for 'fill' is already present. Adding another scale for 'fill', which will replace the existing scale.

2.6.2 Modeling

coeftest(lm_corona_dom_bottom, vcov = vcovHC(lm_corona_dom_bottom, method = "White2", type = "HC0"))

t test of coefficients:

                                             Estimate  Std. Error  t value  Pr(>|t|)    
(Intercept)                                1.7107e+05  3.2306e+04   5.2953 1.199e-07 ***
ac_typenone                               -4.3404e+04  1.9672e+03 -22.0638 < 2.2e-16 ***
ac_typenot_central                        -1.3244e+04  1.5285e+03  -8.6644 < 2.2e-16 ***
patio1                                     7.9510e+03  7.5105e+02  10.5866 < 2.2e-16 ***
school_general1                            1.0340e+04  1.0042e+03  10.2968 < 2.2e-16 ***
photo_count                                9.5789e+02  4.6938e+01  20.4074 < 2.2e-16 ***
pool1                                      9.9372e+03  1.3281e+03   7.4821 7.561e-14 ***
roof_typeother                             2.6563e+03  1.4045e+03   1.8913 0.0585958 .  
roof_typeshingle                           1.9971e+04  1.6039e+03  12.4519 < 2.2e-16 ***
roof_typeslate                             6.1511e+03  8.7277e+03   0.7048 0.4809591    
gas_typenatural                           -1.0026e+05  3.4236e+03 -29.2838 < 2.2e-16 ***
gas_typenone                              -1.3330e+05  2.4175e+03 -55.1396 < 2.2e-16 ***
gas_typepropane                           -1.0165e+05  1.7735e+04  -5.7315 1.007e-08 ***
gas_typeunknown                           -1.3740e+05  2.3153e+03 -59.3429 < 2.2e-16 ***
out_building1                             -5.0020e+03  8.0229e+02  -6.2347 4.601e-10 ***
area_living                                4.3924e+01  6.0025e+00   7.3176 2.603e-13 ***
land_acres                                 3.1769e+03  9.3297e+02   3.4052 0.0006623 ***
appliances1                                2.4716e+04  1.1047e+03  22.3735 < 2.2e-16 ***
garage1                                    1.2482e+04  7.4492e+02  16.7561 < 2.2e-16 ***
property_conditionnew                     -2.2693e+04  6.0403e+03  -3.7569 0.0001725 ***
property_conditionother                   -2.0434e+04  9.2182e+02 -22.1671 < 2.2e-16 ***
energy_efficient1                          1.3977e+04  8.1089e+02  17.2373 < 2.2e-16 ***
exterior_typemetal                         1.0138e+02  2.3235e+03   0.0436 0.9651991    
exterior_typeother                         1.0975e+04  1.0409e+03  10.5441 < 2.2e-16 ***
exterior_typevinyl                         5.0221e+03  1.0863e+03   4.6229 3.803e-06 ***
exterior_typewood                          2.6632e+03  1.7144e+03   1.5535 0.1203185    
exterior_featurescourtyard                 4.1749e+04  1.4307e+04   2.9182 0.0035241 ** 
exterior_featuresfence                    -1.5150e+04  4.9367e+03  -3.0688 0.0021514 ** 
exterior_featuresnone                     -8.7504e+03  4.9532e+03  -1.7666 0.0773066 .  
exterior_featuresporch                    -1.5513e+04  5.0065e+03  -3.0985 0.0019472 ** 
exterior_featurestennis_court              1.8387e+04  1.0849e+04   1.6947 0.0901406 .  
fireplace1                                 1.1927e+04  8.0873e+02  14.7472 < 2.2e-16 ***
foundation_typeslab                        1.3986e+04  1.2569e+03  11.1274 < 2.2e-16 ***
foundation_typeunspecified                 8.1005e+03  1.3966e+03   5.8003 6.702e-09 ***
beds_total1                               -3.0210e+04  2.6492e+04  -1.1404 0.2541445    
beds_total2                               -3.8584e+04  2.6387e+04  -1.4622 0.1436894    
beds_total3                               -3.8581e+04  2.6423e+04  -1.4601 0.1442628    
beds_total4                               -3.4451e+04  2.6453e+04  -1.3024 0.1928085    
beds_total5                               -4.9987e+04  2.6861e+04  -1.8609 0.0627692 .  
bath_full1                                -3.1190e+04  2.3436e+04  -1.3308 0.1832546    
bath_full2                                -9.6994e+03  2.3424e+04  -0.4141 0.6788250    
bath_full3                                 1.2111e+04  2.3511e+04   0.5151 0.6064690    
bath_full4                                 8.9598e+03  2.9138e+04   0.3075 0.7584715    
bath_full6                                -1.3950e+04  2.4076e+04  -0.5794 0.5623287    
bath_half1                                 1.1111e+04  1.0808e+03  10.2800 < 2.2e-16 ***
bath_half2                                 3.1065e+04  6.8469e+03   4.5371 5.730e-06 ***
bath_half3                                 5.8056e+04  1.1205e+04   5.1814 2.220e-07 ***
bath_half4                                 8.7758e+04  3.1894e+03  27.5153 < 2.2e-16 ***
bath_half5                                -5.6301e+04  2.8378e+04  -1.9839 0.0472744 *  
age                                       -1.9126e+03  7.9668e+01 -24.0072 < 2.2e-16 ***
sold_date                                  2.6907e-01  4.6365e-01   0.5803 0.5616918    
sewer_typeseptic                          -5.6527e+03  1.4179e+03  -3.9865 6.724e-05 ***
sewer_typeunspecified                     -4.2515e+03  7.3564e+02  -5.7793 7.592e-09 ***
property_stylenot_mobile                   6.7594e+04  1.7413e+03  38.8190 < 2.2e-16 ***
subdivision1                               3.5261e+03  8.9118e+02   3.9567 7.621e-05 ***
water_typewell                             1.3123e+03  3.9007e+03   0.3364 0.7365576    
waterfront1                                1.9832e+04  1.4576e+03  13.6066 < 2.2e-16 ***
bottom25_dom1                              1.3495e+04  9.6775e+02  13.9442 < 2.2e-16 ***
age_2                                      1.7151e+01  1.1195e+00  15.3204 < 2.2e-16 ***
area_living_2                              5.3725e-03  1.7186e-03   3.1262 0.0017730 ** 
data_factor$infections_3mma                1.0086e+01  7.3559e-01  13.7116 < 2.2e-16 ***
bottom25_dom1:data_factor$infections_3mma -2.1852e+00  8.9042e-01  -2.4541 0.0141291 *  
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

2.7 Corona on City

2.7.1 Visualization

# Distribution: Total 
ggplot(data = data_factor, aes(x = sold_price)) +
    geom_density(mapping = aes(fill = low, alpha = 0.5, position = "identity")) +
    ggtitle("Price Distributions of All Properties") +
    theme(legend.position = "none") +
    xlab("Sold Price") +
    ylab("Density") +
    scale_fill_manual(values = c(very_low))

Warning: Ignoring unknown aesthetics: position

# Distribution: City vs non-city

# Conditional Mean: City vs Rural
library(plyr)
city_limits_mean_data <- ddply(data_factor, "city_limits", summarise, city_limits_mean = mean(sold_price, na.rm = TRUE))

ggplot(data = data_factor, aes(x = sold_price, fill = city_limits)) +
    geom_density(alpha = 0.5, position = "identity") +
    ggtitle("Price Distributions of City vs Rural") +
    geom_vline(data = city_limits_mean_data, aes(xintercept = city_limits_mean_data[2,2]), linetype="dashed", size= 0.5, color = med, alpha = 0.8) +
    geom_vline(data = city_limits_mean_data, aes(xintercept = city_limits_mean_data[1,2]), linetype="dashed", size= 0.5, alpha = 0.8, color = very_low) +
    scale_fill_manual(values = c(very_low, med), 
                      name = "Infection Period",
                      labels = c("Rural", "City")) +
    xlab("Sold Price") +
    ylab("Density") 

# Conditional Mean: City pre vs post corona
library(plyr)
city_limits_mean_data <- ddply(subset(data_factor, data_factor$city_limits ==1 ), "infections_period", summarise, city_limits_mean = mean(sold_price, na.rm = TRUE))

# Distribution: Just City

# Conditional Mean: City pre vs post corona
library(plyr)
city_limits_mean_data <- ddply(subset(data_factor, data_factor$city_limits ==1 ), "infections_period", summarise, city_limits_mean = mean(sold_price, na.rm = TRUE))

ggplot(data = subset(data_factor, data_factor$city_limits ==1), aes(x = sold_price)) +
    geom_density(alpha = 0.5, position = "identity", fill = very_low) +
    ggtitle("Price Distribution of Properties in City Limits") +
    geom_vline(aes(xintercept = mean(city_limits)), linetype="dashed", size= 0.4, alpha = 0.5) +
    xlab("Sold Price") +
    ylab("Density")

Warning in mean.default(city_limits) :
  argument is not numeric or logical: returning NA
Warning: Removed 23399 rows containing missing values (geom_vline).

# Distribution: Infection
ggplot(data = subset(data_factor, data_factor$city_limits ==1), aes(x = sold_price, fill = infections_period)) +
    geom_density(alpha = 0.5, position = "identity") +
    ggtitle("Price Distributions of Properties in City Limits") +
    geom_vline(data = city_limits_mean_data, aes(xintercept = city_limits_mean_data[2,2]), linetype="dashed", size= 0.5, color = med, alpha = 0.8) +
    geom_vline(data = city_limits_mean_data, aes(xintercept = city_limits_mean_data[1,2]), linetype="dashed", size= 0.5, alpha = 0.8, color = very_low) +
    scale_fill_manual(values = c(very_low, med), 
                      name = "Infection Period",
                      labels = c("Pre", "Post")) +
    xlab("Sold Price") +
    ylab("Density") 

#city_limits on Infections
ggplot(data_factor, aes(x = city_limits, y = sold_price, fill = infections_period)) +
    geom_violin(alpha = 0.5) +
    geom_boxplot(width=0.1, alpha = 0.9) +
    scale_fill_viridis(discrete = TRUE, alpha=0.6, option="D") +
    #coord_flip() +
    theme_ipsum() +
    theme(
      legend.position="none",
      plot.title = element_text(size=14)) +
    ggtitle("Comparison of Price: City Limts and Pre vs. Post Corona") +
    xlab("City Limits and Infection Period") +
    ylab("Sold Price") +
    scale_fill_manual(values = c(very_low, med), 
                      name = "City Limits",
                      labels = c("Pre", "Post"))

Scale for 'fill' is already present. Adding another scale for 'fill', which will replace the existing scale.

2.7.2 Modeling

# Testing Corona, City Limits
lm_corona_city <- lm(sold_price ~ . 
               
                       # test variable(s)                    
                       + data_factor$infections_3mma + data_factor$city_limits 
                       + data_factor$infections_3mma*data_factor$city_limits
                       
                       
                       ,data = data_factor_core_clean)
coeftest(lm_corona_city, vcov = vcovHC(lm_corona_city, method = "White2", type = "HC0"))

t test of coefficients:

                                                        Estimate  Std. Error  t value  Pr(>|t|)    
(Intercept)                                           1.6039e+05  3.3079e+04   4.8487 1.251e-06 ***
property_typeDUP                                     -5.1534e+04  1.4834e+04  -3.4740 0.0005136 ***
property_typeOTH                                      1.8189e+04  1.2173e+04   1.4942 0.1351256    
property_typePAT                                      1.6010e+04  5.5366e+03   2.8916 0.0038360 ** 
property_typeSGL                                      2.1938e+04  2.6434e+03   8.2993 < 2.2e-16 ***
property_typeTNH                                     -3.7855e+03  3.1523e+03  -1.2009 0.2298107    
ac_typenone                                          -4.4470e+04  1.9522e+03 -22.7800 < 2.2e-16 ***
ac_typenot_central                                   -1.3891e+04  1.5271e+03  -9.0964 < 2.2e-16 ***
patio1                                                8.1774e+03  7.4902e+02  10.9175 < 2.2e-16 ***
school_general1                                       1.0584e+04  1.0117e+03  10.4613 < 2.2e-16 ***
photo_count                                           9.5668e+02  4.6649e+01  20.5079 < 2.2e-16 ***
pool1                                                 1.2070e+04  1.3348e+03   9.0422 < 2.2e-16 ***
roof_typeother                                        3.4121e+03  1.4035e+03   2.4312 0.0150581 *  
roof_typeshingle                                      2.0811e+04  1.6018e+03  12.9922 < 2.2e-16 ***
roof_typeslate                                        6.5989e+03  8.7749e+03   0.7520 0.4520463    
gas_typenatural                                      -9.6132e+04  3.4602e+03 -27.7820 < 2.2e-16 ***
gas_typenone                                         -1.2948e+05  2.4153e+03 -53.6095 < 2.2e-16 ***
gas_typepropane                                      -9.5196e+04  1.8080e+04  -5.2651 1.413e-07 ***
gas_typeunknown                                      -1.3466e+05  2.3010e+03 -58.5230 < 2.2e-16 ***
out_building1                                        -5.8765e+03  7.9985e+02  -7.3470 2.091e-13 ***
area_living                                           4.3348e+01  5.9670e+00   7.2647 3.852e-13 ***
land_acres                                            2.6636e+03  9.2703e+02   2.8732 0.0040666 ** 
appliances1                                           2.4899e+04  1.1024e+03  22.5853 < 2.2e-16 ***
garage1                                               1.1877e+04  7.4397e+02  15.9649 < 2.2e-16 ***
property_conditionnew                                -1.9899e+04  5.8217e+03  -3.4181 0.0006317 ***
property_conditionother                              -2.1005e+04  9.2539e+02 -22.6983 < 2.2e-16 ***
energy_efficient1                                     1.3781e+04  8.0737e+02  17.0684 < 2.2e-16 ***
exterior_typemetal                                    3.9234e+01  2.3064e+03   0.0170 0.9864279    
exterior_typeother                                    1.0902e+04  1.0367e+03  10.5161 < 2.2e-16 ***
exterior_typevinyl                                    4.7668e+03  1.0820e+03   4.4055 1.060e-05 ***
exterior_typewood                                     2.5348e+03  1.7040e+03   1.4875 0.1368846    
exterior_featurescourtyard                            3.5349e+04  1.4298e+04   2.4723 0.0134301 *  
exterior_featuresfence                               -2.4675e+04  4.8520e+03  -5.0855 3.693e-07 ***
exterior_featuresnone                                -1.7867e+04  4.8599e+03  -3.6765 0.0002370 ***
exterior_featuresporch                               -2.4614e+04  4.9140e+03  -5.0088 5.514e-07 ***
exterior_featurestennis_court                         1.0451e+04  1.0610e+04   0.9849 0.3246603    
fireplace1                                            1.1915e+04  8.0574e+02  14.7875 < 2.2e-16 ***
foundation_typeslab                                   1.5339e+04  1.2542e+03  12.2301 < 2.2e-16 ***
foundation_typeunspecified                            8.7877e+03  1.3912e+03   6.3165 2.722e-10 ***
beds_total1                                          -3.1675e+04  2.4993e+04  -1.2674 0.2050325    
beds_total2                                          -4.3482e+04  2.4904e+04  -1.7460 0.0808248 .  
beds_total3                                          -4.9030e+04  2.4965e+04  -1.9640 0.0495468 *  
beds_total4                                          -4.5468e+04  2.5001e+04  -1.8186 0.0689778 .  
beds_total5                                          -6.0599e+04  2.5432e+04  -2.3828 0.0171878 *  
bath_full1                                           -3.1091e+04  2.4048e+04  -1.2929 0.1960699    
bath_full2                                           -7.6113e+03  2.4040e+04  -0.3166 0.7515426    
bath_full3                                            1.5205e+04  2.4126e+04   0.6302 0.5285382    
bath_full4                                            1.1559e+04  2.9693e+04   0.3893 0.6970565    
bath_full6                                            1.8900e+04  2.4828e+04   0.7613 0.4465139    
bath_half1                                            1.2548e+04  1.0814e+03  11.6033 < 2.2e-16 ***
bath_half2                                            3.1437e+04  6.9326e+03   4.5347 5.795e-06 ***
bath_half3                                            5.7186e+04  1.2061e+04   4.7416 2.133e-06 ***
bath_half4                                            8.3107e+04  3.1313e+03  26.5408 < 2.2e-16 ***
bath_half5                                           -5.6771e+04  2.8810e+04  -1.9705 0.0487890 *  
age                                                  -1.9113e+03  7.9957e+01 -23.9041 < 2.2e-16 ***
dom                                                  -2.2953e+01  6.3300e+00  -3.6260 0.0002884 ***
sold_date                                             2.7729e-01  4.5860e-01   0.6046 0.5454273    
sewer_typeseptic                                     -5.6058e+03  1.4254e+03  -3.9327 8.423e-05 ***
sewer_typeunspecified                                -4.1920e+03  7.3928e+02  -5.6703 1.442e-08 ***
property_stylenot_mobile                              6.8072e+04  1.7313e+03  39.3185 < 2.2e-16 ***
subdivision1                                          3.9045e+03  8.8841e+02   4.3949 1.113e-05 ***
water_typewell                                        4.8366e+03  4.0156e+03   1.2044 0.2284313    
waterfront1                                           1.9906e+04  1.4521e+03  13.7081 < 2.2e-16 ***
bottom25_dom1                                         1.0916e+04  9.6228e+02  11.3434 < 2.2e-16 ***
age_2                                                 1.6963e+01  1.1259e+00  15.0670 < 2.2e-16 ***
area_living_2                                         5.2117e-03  1.7073e-03   3.0525 0.0022715 ** 
data_factor$infections_3mma                           4.2364e+00  1.6257e+00   2.6059 0.0091701 ** 
data_factor$city_limits1                              6.3803e+03  2.1725e+03   2.9369 0.0033181 ** 
data_factor$infections_3mma:data_factor$city_limits1  5.1668e+00  1.6313e+00   3.1673 0.0015405 ** 
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

# Table output
tab_model(lm_corona_city, 
          auto.label = TRUE, 
          collapse.ci = TRUE,
          terms = c("(Intercept)",
                    "data_factor$infections_3mma",
                    "data_factor$city_limits1",
                    "data_factor$infections_3mma:data_factor$city_limits1"))

3. Playground

3.1 Index creation

3.2 Playing with Maps

# packages
require(ggplot2)
install.packages("ggmap")
require(maps)
install.packages(Geoc)



#Basic Map
LA <- map_data("state", region="louisiana")
ggplot(LA, aes(x=long, y=lat))+geom_polygon()


# data
salesCalls <- data.frame(State=rep("louisiana",5), 
                             City=c("Baton Rouge","New Orleans", "Shreveport",       "Lafayette", "Mandeville"),
                             Calls=c(10,5,8,13,2))

salesCalls <- cbind(geocode(as.character(salesCalls$City)), salesCalls)



?cbind

ggplot(LA, aes(x=long, y=lat)) +
  geom_polygon() +
  coord_map() +
  geom_point(data=salesCalls, aes(x=lon, y=lat, size=Calls), color="orange")

3.3 Reduction in Dimensionality

library(boot) # K-fold
library(leaps) # Subset 
library(glmnet) #glmnet() is the main function in the glmnet package (must pass in an x matrix as well as a y vector)

# Set x-y definitions for glmnet package 
x <- model.matrix(sold_price ~ . ,data = data_factor_core_clean)[, -1]

y <- data_factor_core_clean$sold_price[1:24653] # Manually restricted due rows not matching with x 'x' for an unknown reason

# General grid
grid <- exp(seq(10, -65, length = 101)) #grid of values from exp(10) [null model] to exp(-15) [least squares]

#Lasso
set.seed(1)
cv.out <- cv.glmnet(x, y, alpha = 1, lambda = grid, nfolds = 10) #lasso
plot(cv.out)

# Base decision
bestlam <- cv.out$lambda.min; bestlam; log(bestlam)
out <- cv.out$glmnet.fit
lasso.coef <- predict(out, type = "coefficients", s = bestlam); lasso.coef; lasso.coef[lasso.coef != 0]
sum(abs(lasso.coef[1:31])) #l1 norm

# +1se decision
bestlam2 <- cv.out$lambda.1se; bestlam2; log(bestlam2)
lasso.coef2 <- predict(out, type = "coefficients", s = bestlam2); lasso.coef2; lasso.coef2[lasso.coef2 != 0]
sum(abs(lasso.coef2[2:31])) #l1 norm

3.4 Basic 3D Graphing

kd <- with(MASS::geyser, MASS::kde2d(sold_price, infections_3mma, n = 50))

fig <- plot_ly(x = kd$x, y = kd$y, z = kd$z) %>% add_surface()

fig

3.4 Descriptive Stats

# Correlation Matrix heatmap
# Get numeric variable

data_factor$bath_full < as.numeric(data_factor$bath_full)
num_vars <- data_factor %>% dplyr::select(where(is.numeric))
num_vars <- subset(num_vars, select = -c(top50_sold_price))

# Corr matrix
cormat <- round(cor(num_vars),2)
head(cormat)

melted_cormat <- melt(cormat)
head(melted_cormat)

ggplot(data = melted_cormat, aes(x=Var1, y=Var2, fill = value)) + 
   geom_tile() +
   scale_fill_gradient2(low = very_low, 
                        high = high, 
                        mid = med, 
                        midpoint = 0, 
                        limit = c(-1,1), 
                        space = "Lab", 
                        name="Correlation") +
   theme_minimal() + 
   theme(axis.text.x = element_text(angle = 45, vjust = 1, size = 10, hjust = 1, color = "#2E2E2E"),
         axis.text.y = element_text(angle = 0, vjust = 1, size = 10, hjust = 1, color = "#2E2E2E")) +
   coord_fixed() +
   labs(title = "Correlation Matrix",
        x = "",
        y = "")

# Distribution: Total
a <- ggplot(data_factor, aes(x = sold_price/1000)) +
    geom_density(alpha = 0.5, position = "identity", fill = very_low) +
    xlab("Sold Price") +
    ylab("") + 
    theme(axis.text.y=element_blank(), 
          axis.ticks.y=element_blank(),
          text = element_text(size=10)) 

b <- ggplot(data_factor, aes(x = list_price/1000)) +
    geom_density(alpha = 0.5, position = "identity", fill = very_low) +
    xlab("List Price") +
    ylab("") + 
    theme(axis.text.y=element_blank(), 
          axis.ticks.y=element_blank(),
          text = element_text(size=10)) 


c <- ggplot(data_factor, aes(x = area_living)) +
    geom_density(alpha = 0.5, position = "identity", fill = very_low) +
    xlab("Living Area") +
    ylab("") + 
    theme(axis.text.y=element_blank(), 
          axis.ticks.y=element_blank(),
          text = element_text(size=10)) 

d <- ggplot(data_factor, aes(x = land_acres)) +
    geom_density(alpha = 0.5, position = "identity", fill = very_low) +
    xlab("Land in Acres") +
    ylab("") + 
    theme(axis.text.y=element_blank(), 
          axis.ticks.y=element_blank(),
          text = element_text(size=10)) 

e <- ggplot(data_factor, aes(x = area_total)) +
    geom_density(alpha = 0.5, position = "identity", fill = very_low) +
    xlab("Total Area") +
    ylab("") + 
    theme(axis.text.y=element_blank(), 
          axis.ticks.y=element_blank(),
          text = element_text(size=10)) 

f <- ggplot(data_factor, aes(x = age)) +
    geom_density(alpha = 0.5, position = "identity", fill = very_low) +
    xlab("Age") +
    ylab("") + 
    theme(axis.text.y=element_blank(), 
          axis.ticks.y=element_blank(),
          text = element_text(size=10)) 

g <- ggplot(data_factor, aes(x = dom)) +
    geom_density(alpha = 0.5, position = "identity", fill = very_low) +
    xlab("DOM") +
    ylab("") + 
    theme(axis.text.y=element_blank(), 
          axis.ticks.y=element_blank(),
          text = element_text(size=10)) 

data_factor$sold_date <- as.Date(data_factor$sold_date)
str(data_factor)
h <- ggplot(data_factor, aes(x = sold_date)) +
    geom_density(alpha = 0.5, position = "identity", fill = very_low) +
    xlab("Sold Date") +
    ylab("") + 
    theme(axis.text.y=element_blank(), 
          axis.ticks.y=element_blank(),
          text = element_text(size=10)) +
    scale_x_date(date_labels = "%Y")

i <- ggplot(data = subset(data_factor, data_factor$infections_daily > 1), aes(x = infections_daily)) +
    geom_density(alpha = 0.5, position = "identity", fill = very_low) +
    xlab("Infections Daily") +
    ylab("") + 
    theme(axis.text.y=element_blank(), 
          axis.ticks.y=element_blank(),
          text = element_text(size=10)) 

data_factor$beds_total <- as.numeric(data_factor$beds_total)
j <- ggplot(data_factor, aes(x=beds_total)) +
    geom_density(alpha = 0.5, position = "identity", fill = very_low) +
    scale_fill_manual(values=c(very_low)) +
    xlab("Number of Bedrooms") +
    ylab("") + 
    theme(axis.text.y=element_blank(), 
          axis.ticks.y=element_blank(),
          text = element_text(size=10)) 

data_factor$bath_full <- as.numeric(data_factor$bath_full)
k <- ggplot(data_factor, aes(x=bath_full)) +
    geom_density(alpha = 0.5, position = "identity", fill = very_low) +
    scale_fill_manual(values=c(very_low)) +
    xlab("Number of Full Bathrooms") +
    ylab("") + 
    theme(axis.text.y=element_blank(), 
          axis.ticks.y=element_blank(),
          text = element_text(size=10)) 

data_factor$bath_half <- as.numeric(data_factor$bath_half)
l <- ggplot(data_factor, aes(x=bath_half)) +
    geom_density(alpha = 0.5, position = "identity", fill = very_low) +
    scale_fill_manual(values=c(very_low)) +
    xlab("Number of Half Bathrooms") +
    ylab("") + 
    theme(axis.text.y=element_blank(), 
          axis.ticks.y=element_blank(),
          text = element_text(size=10)) 

gridExtra::grid.arrange(a,b,c,d,e,f,g,h,i,j,k,l, nrow =4, ncol = 3)

3.4 Simple UCLA Case

lm_ucla <- lm(sold_price ~ pool + infections_period + pool*infections_period, data = data_factor)
summ(lm_ucla)

# load package
library(sjPlot)
library(sjmisc)
library(sjlabelled)

tab_model(lm_ucla)

3.5 Imagine Combinations

3.6 Tables

end of document